The proposed study will employ theoretical computational methods to investigate the molecular specificity of enzymes on a detailed atomic level. The specific target enzyme/substrate system is the enzyme bovine trypsin and a series of substituted benzamidine inhibitors. The project will consist of three phases directed towards gaining a comprehensive picture of enzyme specificity, from the diffusional approach of the ligand-enzyme pair to the formation of a non-covalent complex of the inhibitor at the specificity site of the enzyme. In Phase I an empirical potential energy function developed for modeling proteins on a detailed atomic level will be used to determine optimum structures and binding energies for a series of substituted benzamidine inhibitors bound to trypsin, and to evaluate binding energy and structural differences in terms of hydrogen bonding, electrostatic, van der Waals and other interactions. Approximate binding constant differences between various inhibitors calculated by this static approach will be compared with experimental binding constant data to determine the validity of studies which ignore explicit solvent and entropy effects. In Phase II the molecular dynamics approach, which simulates the thermal atomic motions in a protein/inhibitor complex, will be used to calculate true thermodynamic binding constants for trypsin bound to the series of inhibitors. These calculations will take into account the detailed interactions with solvent molecules and entropic contributions. In Phase III a Brownian dynamics approach will be used to study the dynamics of the diffusional encounter stage of enzyme/ligand reactions. This approach is based on a stochastic simulation of motion of enzyme/ligand pairs in a solvent medium. This motion is influenced by a complicated range of intermolecular interactions. The proposed studies will demonstrate the feasibility of using model calculations to determine the affinity of enzymes for various substrates and to determine what features must be included in order to understand enzyme specificity. This will ultimately be of practical use in predicting the biological activity of synthetic drugs or enzymes modified to carry out preselected tasks.